AI hears us, LinkedIn polices us
The AI wave is here, and la gente are watching who gets left out. From hearing glasses to LinkedIn’s new rules, the hustle shifts. Your data stays yours. The platform changes. Esto te toca.
Legato Frames: AI hearing glasses for $999
Legato Frames just dropped. The startup’s AI hearing-assistance tech is now built into eyewear. It targets adults with up to moderate hearing loss. Starting price is $999.
This is not a hearing aid in a box. It’s glasses. The AI runs on-device, which means it processes sound right there in the frame. No cloud dependency. That’s a meaningful shift for privacy and reliability.
The target is clear: people who want something that looks like regular glasses but helps them hear in noisy environments. Think family dinners, street markets, crowded buses. The kind of settings where the comidad gets lost in the noise.
Legato is positioning itself as a consumer product, not a medical device. That matters. It means the design, the fit, the style — all the things that make eyewear something people actually wear. The $999 starting price puts it in the premium eyewear lane, not the budget hearing-aid lane.
The on-device AI angle is the real story here. Hearing-assistance tech has existed for years. What’s new is the ability to run the noise-cancellation and amplification locally, in real time, without sending your audio to a server. For the comunidad, that’s a practical win. Your conversations stay yours.
No clinical trial data is cited in the source. No user testimonials. No independent review. It’s a launch announcement. Take the specs and the price tag at face value. The real test is whether people actually wear them and whether the AI holds up in the wild.
Why this matters for us: When hearing tech finally looks and feels like something our abuelos would actually wear, the stigma drops — and that’s when the whole familia benefits.
AI companies going public — who else rides the wave?
The big AI firms are lining up for public markets. Startups are trying to ride that SpaceX IPO wave, looking for the same kind of momentum. The question is who else gets a seat along the way.
The rush to go public isn't just about the headline names. It ripples through the people behind them — the engineers, the founders, the small teams that built the infrastructure. When a big AI company lists, the market pays attention, and the money follows. That doesn't mean everyone at the table eats.
There's a pattern here that the comunidad knows well. You see the cousin who runs the side business, the auntie on Facebook, the folks who keep the lights on in the small shop. When the big wave hits, they're the ones who feel it first — sometimes as opportunity, sometimes as a squeeze. The same way a big developer moving into the neighborhood changes rent, a big AI IPO changes who gets funded, who gets hired, who gets left out.
The race to go public is happening now. The companies are talking to investors, setting dates, preparing filings. The startups are watching, calculating, wondering if they can hitch a ride on the momentum. It's not a guaranteed ticket. The wave doesn't carry everyone.
Why this matters for us: When the big AI companies go public, the money, the jobs, and the attention shift — and that affects the small businesses, the side hustles, and the families who depend on that local economy.
RAG: How AI finally stops making stuff up
You know that cousin who tells you he can fix your car, but he’s been watching YouTube for 20 minutes and doesn’t know the difference between a spark plug and a tire valve? That’s an AI without RAG. It’s confident, it speaks Spanish, and it’s lying to you.
RAG is the fix. It stands for Retrieval-Augmented Generation. Here’s what that means in plain English: before the AI writes a sentence, it first goes to a library. It pulls up real documents — PDFs, websites, databases — and reads them. Then it writes its answer based on what it just read, not on what it dreamed up.
Think of it like this: you’re at the barbershop. Your abuelo asks for the recipe for your tía’s mole. A regular AI (without RAG) would guess. It’d say “chocolate, chiles, and hope.” But RAG? RAG is the uncle who actually goes to the kitchen, opens the old notebook, reads the recipe out loud, and then tells you what it says. He might be slow. He might say “let me check.” But he’s not making it up.
Why companies care: it stops hallucinations. Your AI now has a source it can point to. “I got this from page 42 of the manual.” That’s huge for health, law, and any job where getting it wrong costs money.
Why this matters for us: most of the tools we’ll use at work — customer service bots, HR systems, compliance checks — will be built on RAG. If you know how it works, you’ll spot when the system is guessing vs. when it’s reading the actual rules. Ask your boss: “Where is this pulling from?” That one question separates the real deal from the cousin who’s been watching YouTube.
The ball doesn’t know the score. It only knows the forces acting on it.
— wired.com
#the-physics-of-the-curl-why-soccer-shots-bend-midair-0f4da9Siri AI on the Mac is finally waking up
You know the feeling. You turn off the assistant, ignore the prompts, and move on. That’s how most of us treated Siri on the Mac. And Apple Intelligence felt like a placeholder. Fruitless. No point. Just another feature we scrolled past.
But now the new Siri AI is here on macOS 27 Golden Gate, and it’s making people sit up straight. Not because it’s perfect. Because it’s not useless anymore.
I’ve been testing it for 24 hours on a dev beta. It’s early. The indexing is still running. It’s not done. But it’s different. The gap between “la migra app” and something that actually understands what I’m typing is closing.
Obsidian AI: Your data stays put. The brain stays yours.
You have files the law won't let leave the building. IEPs. Medical records. Legal discovery. City HR files. The staff still needs AI — drafting, summarizing, translating, looking things up. But most "private AI" still phones home. And most fully-local stacks are a research project, not something you can just drop on the network and start using.
Obsidian AI is the turn-key appliance. GPU, model, agent runtime, voice, and a hardened admin console — all on your own network. No outbound calls. Same toolbox surface as BFTS Chat, but the data and the brain never leave the room. If the rule says "no cloud," this respects that rule.
For school districts, clinics, county legal, and anyone under HIPAA, FERPA, or CJIS constraints — this is the difference between a compliance risk and a tool that actually works.
Why this matters for us: la gente in our communities trust the system with their kids' records, their medical history, their legal files. When the tech sends that data out, it's not a glitch — it's a betrayal. Obsidian AI gives those orgs a real option that keeps la gente's data where it belongs.
https://brownforces.io/solutions
X-Men '97 season 2 vs. Masters of the Universe: nostalgia done right
Marvel and Mattel are both cashing in on the same well in 2026: the animated heroes we watched as kids. Masters of the Universe just dropped a live-action He-Man movie. X-Men '97 season 2 is about to launch on Disney Plus, and the mutants are heading into an apocalyptic future.
Both projects were clearly made by people who love the source material. Both are stuffed with nerdy Easter eggs meant to get the hardcore fans hyped. They are both nostalgia plays.
But there is a difference. The article suggests X-Men '97 is pulling ahead where Masters of the Universe is missing something.
LinkedIn's Creator Marketplace: New rules for Brown creators
LinkedIn is rolling out its Creator Marketplace, and the new rules are already reshaping how creators sell, partner, and get paid. The platform is calling it an "ecosystem." We'll call it what it is: a new gate for creators who built audiences outside the corporate ladder.
The marketplace lets creators list their services — consulting, coaching, content strategy, workshops — and connect directly with brands. No more cold DMs. No more guessing who's actually hiring. Brands can filter by niche, audience size, and engagement rate. Creators set their own rates. LinkedIn takes a cut on transactions.
But here's what the fine print doesn't highlight: the algorithm still decides who gets visibility. If you're not posting daily, you're invisible. If your last post was three weeks ago, your listing gets buried. This isn't a level playing field — it's a performance tax.
For Brown creators, the question isn't whether the marketplace exists. It's whether the new rules favor the same people who've always had the connections. The algorithm rewards consistency. And consistency is expensive when you're juggling a side hustle, a family, and a job that doesn't care about your content.
The marketplace also introduces new disclosure rules. Creators must label sponsored content. Brands must approve posts before they go live. This isn't about protecting consumers — it's about giving brands more control over the narrative. If your post doesn't pass their review, it doesn't get published. Period.
Some creators are already complaining. The payout structure is unclear. The dispute resolution process is opaque. And the platform's own terms of service give LinkedIn the right to change the rules without notice. That's not a partnership. That's a landlord.
The real test will be whether this marketplace creates real opportunities for creators who don't fit the traditional mold — the ones who built audiences in Spanish, in neighborhoods, in communities that LinkedIn's algorithm has never learned to value.
Why this matters for us: If the new rules favor the same people who've always had the connections, then this marketplace isn't an opportunity — it's another gatekeeper wearing a LinkedIn badge.
Loewe's TikTok play: why the brand's going viral in the feed
Loewe is making a move on TikTok, and it’s not just another brand trying to look cool. The strategy is built around short-form video content that leans into the platform’s native language — humor, trends, and a bit of chaos. This isn’t a traditional ad campaign. It’s content that feels like it belongs in your feed, not a break from it.
The approach is simple: show up, be funny, don’t overthink it. The brand is letting creators lead, which means the content feels less polished and more human. That’s the point. It’s about building a relationship with the audience, not just selling a product.
K8s now routes LLM calls like a dispatcher
Datadog shipped an extension for Kubernetes that acts as an LLM routing layer. You point it at your cluster, feed it a list of models — open-source, hosted, whatever — and it decides which one gets the call.
It tracks latency, cost, and error rates per model. If a model starts lagging or burning through tokens, the router can switch traffic to the next one on the fly. No app changes needed.
Think of it like a cousin running a side hustle. You tell the cousin what you need, they check who's got the fastest response right now, and they hand it off. That's the router.
The extension plugs into existing K8s deployments. You don't rewrite your app to talk to it. You just tell the cluster: "route these LLM calls through me."
Datadog's pitch is cost control. If you're running multiple models — one for fast answers, one for accuracy, one for cheap volume — this lets you shift traffic based on real-time performance instead of guessing.
It's not a new model. It's a dispatcher sitting in front of the models you already pay for.
Why this matters for us: Most of us are already juggling too many tools and too many subscriptions. A routing layer that cuts costs without making us rewrite our apps? That's the kind of infrastructure that lets small teams compete without burning cash on unused model capacity.
How to pick the right LLM before you build
TLDR DevOps shared a GitHub repo called whichllm, a curated list of open-source large language models ranked by performance. The idea is simple: don't guess which model fits your project — look at the scores.
The repo sorts models by benchmarks like MMLU, HumanEval, and coding tasks. It includes models from Meta, Mistral, Google, and others. There's a quick-start section that shows how to run them locally with tools like Ollama or LM Studio.
The author is a developer who builds side projects. The list is updated regularly. No paywall. No vendor lock-in.
This is the kind of resource the cousin who runs the side business needs.
Product managers: your empathy is your edge
Roman Pichler dropped a new episode on TLDR Product. The topic is emotional intelligence in product management.
He breaks down the idea that EQ isn't a soft skill. It's how you read the room before you ship something nobody wants.
The podcast covers three main points:
1. You need to read the team, not just the product.
2. You need to manage your own frustration when a launch tanks.
3. You need to spot when the founder is running on ego, not data.
Pichler says the best PMs aren't the ones with the sharpest Jira boards. They're the ones who can tell when a stakeholder is lying to themselves. That's the quiet part nobody teaches in the bootcamp.
He also touches on how AI is changing the game. AI can write the specs. It can't tell if the VP is actually on board or just nodding to keep the budget flowing.
The episode is long. 45 minutes. But the core idea is simple: your job isn't to manage tasks. It's to manage people. And people are messy. They lie. They panic. They get defensive when you point out a flaw in their idea.
If you're a Brown PM working in a room where you're the only one who looks like you, this is the episode. Because you already know how to read a room. You've been doing it since you were six. Now you just need to call it what it is: a professional skill.
Why this matters for us: The tech industry keeps selling us the idea that the best PMs are the ones with the biggest portfolios. But the ones who actually ship and keep their jobs are the ones who know when to walk out of a meeting. That's not a soft skill. That's survival.